Applications of Artificial Intelligence in the Smart Grid

The power grid has always been a complex system that instigated the invention and implementation of intelligent approaches for solving problems. An early example of such a challenge is the power flow problem, which requires solving a large system of nonlinear equations. With the introduction of embedded systems at all levels of the power grid, new possibilities and challenges arise. The decarbonization of our energy system comes with the need to predict photovoltaic and wind output, which can be addressed using machine learning and Artificial Neural Networks. In durations where power production is low, a reduction of energy consumption is necessary. Artificial intelligence can be applied here to coordinate the operation of devices while keeping user effort and discomfort to a minimum. The best way to help with the energy transition is, of course, to reduce energy consumption in the first place. With the application of machine learning techniques for Non-Intrusive Load Monitoring, it is possible to provide the users detailed feedback about their devices' energy consumption which motivates and induces energy saving by a more careful usage of energy-hungry devices.